{"id":"07cc5e11-b34f-47e7-9fd2-ca2e9c3f883a","arxiv_id":"2412.10007","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Eigenfunctions of Krein-Feller operators on bounded domains of Riemannian manifolds satisfy the Courant nodal bound and are continuous under mild dimension conditions.","lead":"This paper proves a Courant nodal domain theorem and continuity results for eigenfunctions of Krein-Feller operators, a class of measure-driven Laplacians, on Riemannian manifolds. It extends the authors' earlier R^d results to curved spaces using normal coordinates and Green's function arguments.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The maximum principle proof in Theorem 3.3 uses an invalid Euclidean change of variables in (3.6), so the nodal domain theorem 2.1 currently lacks a valid proof.","rationale":"The paper's central new contribution is the Courant nodal domain theorem for Krein–Feller operators on Riemannian manifolds. The proof of Theorem 2.1 hinges on Theorem 3.3, the maximum principle for μ-subharmonic functions. Reading the proof of Theorem 3.3 in good faith, the essential step is the passage from the manifold to a Euclidean chart in (3.6). This passage is incorrect: it identifies the Riemannian volume measure with Lebesgue measure and the Laplace–Beltrami operator with the Euclidean Laplacian. On a curved manifold these are not the same, and no additional argument is supplied to absorb the metric terms. This is not a disagreement with a conventional result; it is an internal inconsistency in the proof. The reader's weakest assumption pointed at this region; our analysis sharpens it from 'only sketched' to 'faulty.' The continuity theorems (2.2 and 2.3) are argued via the Green operator and do not depend on Theorem 3.3, so they may survive independently. However, the nodal theorem is a central claim and is currently unsupported. A correct maximum principle in divergence form on manifolds, or a more careful chart-dependent version of Lemma 3.2, could repair the argument, but it is not provided. For this reason the reader's conditional verdict remains appropriate: the paper should not be accepted as is, but the flaw is localized and does not by itself falsify the theorem.","tokens_in":26900,"tokens_out":19071,"duration_ms":214080,"concrete_test":"Take M=S^2 with the round metric and a geodesic normal coordinate chart centered at the north pole. Choose any smooth u and a compactly supported smooth v near the pole, then evaluate both sides of (3.6) using dν_{S^2}=√g dx and Δ_{S^2}v = g^{-1/2}∂_i(√g g^{ij}∂_j v). The two integrals will not be equal because √g is not identically 1, demonstrating that the coordinate change in (3.6) is invalid as stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Theorem 3.3, the reduction to the Euclidean maximum principle (Lemma 3.2) is not valid for a general non-flat Riemannian manifold. In the normal coordinate chart φ_i, the Riemannian volume is dν = √g dx and the Laplace–Beltrami operator acts as Δ_g v = g^{ij}∂_i∂_j v + (∂_i(√g g^{ij})/√g)∂_j v, which is not the Euclidean Δ(v∘φ_i^{-1}). Equation (3.6) replaces ∫_{Ui0} uΔv dν by ∫_{φ(Ui0)} (u∘φ^{-1}) Δ(v∘φ^{-1}) dx, silently identifying dν with Lebesgue measure and Δ_g with the Euclidean Laplacian. These differ by terms that are generically nonzero on any curved manifold (e.g., on S^2 with the round metric, √g = 1 + O(r^2) in geodesic normal coordinates). Consequently Lemma 3.2—which applies only to the Euclidean operator with Lebesgue measure—cannot be invoked in the manner presented. Since Theorem 2.1's proof relies on Theorem 3.3 to show that the first eigenfunction has a constant sign and to count nodal domains, the nodal bound is not established as written. The continuity theorems (2.2, 2.3) use a different Green-operator argument and are not directly affected by this defect.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Krein-Feller operators Δ_μ on bounded domains of complete Riemannian manifolds and on compact closed manifolds, where μ is a positive finite Borel measure satisfying dim∞(μ) > d−2. It claims a Courant nodal domain theorem (Theorem 2.1) under an assumption that eigenfunctions are continuous, and two theorems (Theorems 2.2 and 2.3) establishing continuity of eigenfunctions on bounded domains and on compact closed manifolds, respectively. The proofs are built on a maximum principle for μ-subharmonic functions, a Green-operator representation of the inverse of −Δ_μ, and small-ball estimates for the Green function. The paper also contains an additional result on conformal Riemann surfaces (Theorem 8.2) with an example.","tokens_in":27193,"tokens_out":13241,"duration_ms":143633,"significance":"If correct, the main results would be the first Courant nodal bound and eigenfunction regularity statements for Krein-Feller operators on Riemannian manifolds, extending the Euclidean results of [40]. The continuity theorems are natural and potentially useful for nodal set analysis and for generalizing Yau-type conjectures to measure-valued Laplacians. The paper includes a concrete example of a continuous eigenfunction on a domain of the sphere. However, the current manuscript has serious gaps: the maximum principle is not proved as written due to an invalid coordinate change, and the Green-operator sections contain systematic sign errors. These issues affect the proofs of Theorems 2.1–2.3, though they appear to be repairable with additional work.","major_comments":[{"comment":"The proof of Theorem 3.3 is not valid as written. Equation (3.6) replaces the manifold integral ∫_{U0} uΔv dν by the Euclidean integral ∫_{φ(U0)} (u∘φ^{-1}) Δ(v∘φ^{-1}) dx, but in geodesic normal coordinates the Laplace–Beltrami operator is Δ_g v = g^{ij}∂_i∂_j v + (∂_i(√g g^{ij})/√g)∂_j v and the volume form is √g dx, not the Euclidean Laplacian and Lebesgue measure. These differ by generically nonzero terms on any curved manifold, so Lemma 3.2, which applies only to the Euclidean operator with Lebesgue measure, cannot be invoked. In addition, (3.6) has a sign error: combining (2.6) with the identity ∫⟨∇u,∇v⟩dν = −∫uΔv dν gives ∫uΔv dν = −∫f v dµ, not +. Since Theorem 2.1 relies on Theorem 3.3 to show that the first eigenfunction has constant sign and to count nodal domains, the nodal bound is not established as presented.","section":"Theorem 3.3, equation (3.6)"},{"comment":"There is a persistent sign inconsistency in the Green operator argument. In Proposition 5.6 the authors state that G_μ f solves Δu = f μ, and in (5.11) they compute ∫(G_μ f)Δξ dν = ∫ξ f dμ. However, using their own Green function equation (5.5), −ΔG_y = δ_y, the correct integration by parts gives ∫G_y Δξ dν = −ξ(y), hence G_μ f solves Δu = −f μ, not +f μ. The same error appears in (7.12) of Section 7. Consequently the statements of Propositions 5.6 and 7.3, and the derivation of Theorems 5.9 and 7.4, are not correct as written. The final inverse statements may be salvageable after flipping signs in the intermediate equations, but as it stands the proofs of Theorems 2.2 and 2.3 rest on an inconsistent sign convention. Moreover, in the proof of Theorem 2.3 the eigenvalue equation is written as Δ_μ f = λf, whereas in Section 2 eigenfunctions are defined by −Δ_μ u = λu; this needs reconciliation.","section":"Sections 5 and 7, Propositions 5.6, 5.11, 7.3, 7.12"},{"comment":"The uniform small-ball estimates (6.14) and (6.15) are asserted without proof. Condition (5.6) only bounds the full integral ∫Ω G_y(x)dμ(y) uniformly in x; it does not by itself imply that the integral over a small ball B_r(z) tends to zero uniformly in x. For the d ≥ 3 case, such a bound can be derived from the α-regularity of μ and the pointwise singularity of the Green function, but the derivation is not given. For d = 2, a similar uniform estimate for the squared Green function is needed. Since these estimates are the core of the continuity proof in Proposition 6.6, Theorems 2.2 and 2.3 depend on this missing step. The authors should provide the details or a precise reference.","section":"Proposition 6.6, Step 1, equations (6.14)–(6.15)"},{"comment":"Theorem 8.2 invokes [22, Theorem 1] to assert that the Green function pulls back by G^Z_y(x) = G^U_{φ(y)}(φ(x)) under a conformal map between Riemann surfaces. This identity is not generally true for the Laplace–Beltrami operator on a Riemannian surface: a conformal map changes the metric by a conformal factor, and the Green function transforms with additional terms involving the conformal factor. The cited reference is a paper on the method of images for spherical domains and does not establish the general statement needed here. Thus the proof of Theorem 8.2 and the subsequent Example 8.1 are not justified as written. Since this section is not used in the proofs of Theorems 2.1–2.3, this is a separate gap rather than a load-bearing one, but it still affects a stated result.","section":"Section 8, Theorem 8.2"}],"minor_comments":[{"comment":"There are numerous typos, including “Scetion” in Section 2, “manidolds” in the references, “Basis on Theorem 8.2” instead of “Based on Theorem 8.2”, and inconsistent notation such as “∆ μ” and “Δ_μ”. A careful copyedit is needed.","section":"Throughout"},{"comment":"The sentence “By Proposition 5.4, condition (5.6) holds for ν” should refer to μ, not ν. Also, in the inequality the exponent in ∥G_μ f∥_{L^p(Ω)} is written for the Lebesgue measure dν, but the right-hand side is measured with respect to dμ; the distinction should be stated explicitly.","section":"Proposition 5.5, proof"},{"comment":"The last sentence says “Part (a) now follows by using [38, Theorem 2.2]” but the proof is in part (b); this is a typo.","section":"Proof of Proposition 4.1(b)"},{"comment":"The notation G_μ f^2 is ambiguous; it should be written as G_μ(f^2) to avoid confusion with (G_μ f)^2. The same applies to similar expressions elsewhere (e.g., in the proof of Proposition 6.4).","section":"Lemma 6.3 and Proposition 6.4"},{"comment":"The assumption dim∞(ν) > d−2 is automatically satisfied for the Riemannian volume measure on a smooth d-manifold (dim∞(ν) ≥ d by the Bishop–Gromov comparison). This hypothesis can be removed or justified in the text rather than left as an unexplained condition.","section":"Theorem 2.2, hypotheses"},{"comment":"The definition of μ-subharmonic functions would be clearer if it explicitly stated that Δ_μ u is a function in L^2(Ω,μ) and that the inequality is understood pointwise μ-a.e. on Ω.","section":"Definition 3.1"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses an interesting and timely question, and the main theorems are plausible. However, in its current form the maximum principle is not proved because the coordinate chart argument is invalid, and the Green-operator sections have consistent sign errors that affect the proofs of the continuity theorems. These issues are potentially repairable but require substantial rewriting. The repeated reliance on self-cited preprints [38,40] for essential steps (maximum principle, spectral theory, Green estimates) also needs to be made explicit and complete, since the present manuscript does not give enough detail for a reader to verify the reductions. I would recommend a major revision rather than rejection, but the authors should be asked to provide complete proofs of the maximum principle and to correct the sign conventions throughout Sections 5–7."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is not ready as written. The nodal domain theorem (Theorem 2.1) currently has an invalid proof: the maximum principle for µ-subharmonic functions, Theorem 3.3, is proved by pulling the equation back to Euclidean space through normal coordinates and applying the Euclidean lemma 3.2. But the pullback is not Euclidean. In normal coordinates the volume form is √g dx and the Laplace–Beltrami operator is (1/√g)∂_i(√g g^{ij}∂_j), not the flat Laplacian. Equation (3.6) silently identifies dν with dx and Δ_g with Δ_Rd; these differ by terms that do not vanish on a curved manifold. So Lemma 3.2, which is a Euclidean statement with Lebesgue measure, cannot be invoked. Since Theorem 3.3 is the load-bearing step in showing the first eigenfunction has constant sign and in counting nodal domains, Theorem 2.1's proof collapses as written. The stress-test note is right about this.\n\nThe sign inconsistency is also real. In Section 5, (5.5) says −ΔG_y = δ_y, but the computation (5.11) uses ∫ G_y Δξ = ξ(y), which corresponds to ΔG_y = δ_y. The stated theorem (5.9) sets G_µ + h_µ = (−Δ_µ)^{-1}, which is consistent with the intended (5.5) convention, but not with the displayed computation. In Section 7, the Green function is defined with ΔG = δ − 1/vol (no minus), and Theorem 7.4 concludes Δ_µ G_µ f = f, making G_µ the inverse of Δ_µ, not −Δ_µ. In the proof of Theorem 2.3, the authors then use f = λG_µ f to characterize λ-eigenfunctions, which is only correct for −Δ_µ if the sign is adjusted. This is not a cosmetic typo: it feeds directly into the claimed equivalence.\n\nWhat is genuinely new: the extension of the Krein–Feller framework to curved settings, the Green operator decomposition on bounded domains and closed manifolds, the small-ball estimates, and the conformal surface example. The overall strategy follows the earlier R^d papers, which is fine, but the new geometric parts need to be checked carefully. Heavy reliance on self-cited preprints [24, 38, 40] is not itself a flaw, but here the key reduction to [38]'s charts is exactly where the Euclidean mistake sits.\n\nMy guess is the main theorems are true and the paper can be repaired: fix the maximum principle by working with the actual metric, align the sign conventions, and re-check the inverse statements. But as it stands, the nodal theorem lacks a valid proof and the sign issue in Theorem 2.3's proof is unresolved. I would not desk-reject the paper—it deserves a careful referee—but it should not be accepted in this form. I would not cite it yet.","headline":"The nodal theorem's proof rests on a false Euclidean identification in normal coordinates and the Green operator sign flips between sections; the paper is salvageable but not ready as written.","tokens_in":27709,"tokens_out":5850,"would_cite":false,"duration_ms":63308,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J05","35B05","28A80","58C40","35J08"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper establishes a Courant nodal domain theorem for Krein-Feller operators on bounded domains of Riemannian manifolds, conditional on continuity of eigenfunctions, and proves that continuity on smooth bounded domains and on compact…","keywords":["Krein-Feller operator","Riemannian manifold","nodal domain theorem","eigenfunction continuity","Green operator","measure-valued Dirichlet problem","μ-subharmonic maximum principle","lower L∞-dimension"],"falsifier":"Check the maximum-principle transfer on a geodesic ball: exhibit a nonconstant continuous function u with Δ_μu≥0 μ-a.e. that attains its maximum at an interior point of a small geodesic ball in a smooth Riemannian manifold; that would falsify Theorem 3.3 and with it the nodal bound. Alternatively, on a domain where eigenfunctions are known to be continuous, compute the nodal domains of a λ_2-eigenfunction for a measure μ with dim∞(μ)>d-2; finding more than two nodal domains would falsify Theorem 2.1.","tokens_in":26714,"feed_emoji":"📐","tokens_out":7917,"duration_ms":84145,"temperature":0.7,"pith_summary":"This paper proves the first Courant-type nodal domain theorem for Krein-Feller operators—Laplacians defined by a Riemannian metric but acting on L2 with respect to an auxiliary measure μ—on bounded domains of complete Riemannian manifolds. The theorem says that if a λ_n-eigenfunction is continuous, its zero set divides the domain into at most n nodal domains (at most n+1 when the boundary is empty), under the dimension condition dim∞(μ)>d-2. The paper then removes the continuity hypothesis in two cases: on bounded domains with smooth boundary and a Green's function, and on compact connected closed manifolds, all eigenfunctions are continuous. A reader should care because these are the basic tools—nodal counts and pointwise regularity—needed to extend nodal-line and spectral-geometry questions to Laplacians whose volume is carried by an arbitrary measure.","feed_headline":"Krein-Feller eigenfunctions obey Courant nodal bound on manifolds","feed_subtitle":"For sufficiently thick measures, eigenfunctions are continuous, so nodal sets are well defined and bounded by index.","key_machinery":"The central objects are the Krein-Feller operator -Δ_μ, defined as the self-adjoint operator associated with the Dirichlet form E(u,v)=∫_Ω⟨∇u,∇v⟩dν on L2(Ω,μ), and its Green operator G_μf(x)=∫_Ω G_y(x)f(y)dμ(y), built from the Dirichlet Green function of the Laplace-Beltrami operator. The nodal theorem is carried by the maximum principle for μ-subharmonic functions, meaning functions u with Δ_μu≥0 μ-a.e.: a nonconstant continuous such function cannot attain its maximum inside Ω. The paper proves this by pulling the function back through normal coordinate charts and invoking the Euclidean maximum principle. The continuity theorems are carried by the identity (-Δ_μ)^{-1}=G_μ+h_μ, where h_μ is a harmonic correction term, together with estimates showing that G_μ maps the domain of -Δ_μ into bounded continuous functions.","core_discovery":"The central claim is that the Krein-Feller operator -Δ_μ on a bounded domain Ω of a smooth complete Riemannian manifold inherits the Courant nodal bound from the Laplace-Beltrami operator: for each n, a λ_n-eigenfunction that is continuous has at most n nodal domains, or n+1 when ∂Ω=∅. The supporting discovery is that, for d≥2, such eigenfunctions are genuinely continuous whenever the domain has smooth boundary and a Green's function, with both dim∞(μ)>d-2 and the analogous condition for the volume measure, or when the manifold is compact, connected, and closed. The proofs run through a maximum principle for μ-subharmonic functions, obtained by transferring the Euclidean argument through normal coordinate charts, and through an inverse formula expressing -$Δ_μ^{{-1}}$ as the Green operator G_μ plus a harmonic correction.","pith_inferences":["If the chart-patching step in the maximum principle can be made fully explicit and the Euclidean maximum principle holds under the measure Poincaré inequality alone, the nodal bound should extend to arbitrary bounded domains without smoothness, since only compactness and continuity of eigenfunctions are used.","The threshold d-2 appears both in the existence of the operator and in the Green-operator bounds, suggesting that the condition is not merely technical and that examples at exactly dim∞=d-2 may be sharp.","The inverse identity (-Δ_μ)^{-1}=G_μ+h_μ suggests a transfer principle: any regularity theorem proved for the Green potential with respect to μ, such as Hölder continuity under stronger dimension assumptions, would automatically hold for eigenfunctions.","The conformal-surface section points to a broader route: on conformally flat surfaces, Krein-Feller eigenfunctions can be pulled back from the plane, so planar examples and counterexamples transfer directly to manifolds."],"forward_implications":["On a bounded smooth domain with a Green's function, Theorems 2.1 and 2.2 combine: eigenfunctions are continuous, so the Courant nodal bound applies unconditionally to them.","On a compact connected closed manifold, the same combination controls nodal domains of all nonconstant eigenfunctions, with constants playing the role of the λ_0 eigenfunction.","Continuity of eigenfunctions makes it possible to ask manifold versions of nodal-line questions, such as second-eigenfunction nodal geometry and measure-valued analogues of Yau's nodal measure conjecture, for Krein-Feller operators.","The dimension condition dim∞(μ)>d-2 identifies a natural class of measures for which the Krein-Feller spectrum is discrete and eigenfunctions have pointwise meaning.","On conformal Riemann surfaces, eigenfunctions of the pushed-forward measure are eigenfunctions on the original surface, giving concrete continuous eigenfunction examples such as a bounded domain in the sphere."],"supporting_citations":[{"why":"It supplies the normal-coordinate chart construction, the measure Poincaré inequality, and the spectral theory of Δ_μ on manifolds that the proofs rely on.","marker":"[38]"},{"why":"It gives the Euclidean maximum principle and the R^d nodal and continuity results that the paper transfers to the manifold setting.","marker":"[40]"},{"why":"It provides the Green-operator estimates and the inverse-operator argument for Laplacians on measure spaces that the continuity proofs adapt.","marker":"[24]"},{"why":"It sets the weak-solution framework for the Dirichlet problem with density measure that is used to build the Green operator.","marker":"[42]"},{"why":"It supplies existence and boundary regularity of the Dirichlet Green function on Riemannian domains with smooth boundary.","marker":"[2]"},{"why":"It is the manifold Courant nodal theorem for the classical Laplace-Beltrami operator that the result extends.","marker":"[9]"},{"why":"It supplies the baseline Courant nodal-domain theorem and spectral ordering for Laplace-Beltrami operators.","marker":"[43]"}],"fun_headline_variants":["Krein-Feller eigenfunctions respect Courant nodal bound","Nodal bound holds for Krein-Feller eigenfunctions on manifolds","Continuity unlocks Courant bound for Krein-Feller operators","Courant theorem extends to Krein-Feller operators on manifolds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole nodal bound leans on a single transfer step: the maximum principle for μ-subharmonic functions, proved in Euclidean space, still works when pulled back to small curved coordinate patches on the manifold. The paper sketches that step rather than writing out every detail; if the transfer has a hidden gap, the Courant bound falls even if the continuity theorems stand.","fun_headline_variants_meta":{"raw":{"variants":["Krein-Feller eigenfunctions respect Courant nodal bound","Nodal bound holds for Krein-Feller eigenfunctions on manifolds","Continuity unlocks Courant bound for Krein-Feller operators","Courant theorem extends to Krein-Feller operators on manifolds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000186,"raw_usage":{"total_tokens":1288,"prompt_tokens":868,"completion_tokens":420,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":346}},"tokens_in":484,"tokens_out":420,"duration_ms":4351,"temperature":1.0,"reasoning_tokens":346,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T16:28:17.805252+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the maximum-principle transfer on a geodesic ball: exhibit a nonconstant continuous function u with Δ_μu≥0 μ-a.e. that attains its maximum at an interior point of a small geodesic ball in a smooth Riemannian manifold; that would falsify Theorem 3.3 and with it the nodal bound. Alternatively, on a domain where eigenfunctions are known to be continuous, compute the nodal domains of a λ_2-eigenfunction for a measure μ with dim∞(μ)>d-2; finding more than two nodal domains would falsify Theorem 2.1.","supporting_citations":[{"cited_title":"Ngai and L","cited_arxiv_id":null,"evidence_quote":"It supplies the normal-coordinate chart construction, the measure Poincaré inequality, and the spectral theory of Δ_μ on manifolds that the proofs rely on."},{"cited_title":"Nodal sets and continuity of eigenfunctions of Kre\\u{\\i}-Feller operators","cited_arxiv_id":"2411.14173","evidence_quote":"It gives the Euclidean maximum principle and the R^d nodal and continuity results that the paper transfers to the manifold setting."},{"cited_title":"Hu, K.-S","cited_arxiv_id":null,"evidence_quote":"It provides the Green-operator estimates and the inverse-operator argument for Laplacians on measure spaces that the continuity proofs adapt."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It sets the weak-solution framework for the Dirichlet problem with density measure that is used to build the Green operator."},{"cited_title":"Aubin, Nonlinear analysis on manifolds","cited_arxiv_id":null,"evidence_quote":"It supplies existence and boundary regularity of the Dirichlet Green function on Riemannian domains with smooth boundary."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It is the manifold Courant nodal theorem for the classical Laplace-Beltrami operator that the result extends."},{"cited_title":"Schoen and S.-T","cited_arxiv_id":null,"evidence_quote":"It supplies the baseline Courant nodal-domain theorem and spectral ordering for Laplace-Beltrami operators."}],"review_version":1}